J. Neurol. Neurosurg. Psychiat., 1950, 13, 36.

DEGENERATION OF THE PRIMARY AND SECONDARY SENSORYNEURONES AFTER TRIGEMINAL INJECTION

BY

JOHN PENMAN and MARION C. SMITH*

This paper comprises a clinical and pathologicalreport of a case of tic douloureux, in which deathoccurred three and a half months after a successfulinjection of the right trigeminal nerve.Though this form of injection has been practised

for nearly 40 years, there is still dispute about theexact site of action of the alcohol, the changes whichit produces in the trigeminal pathway, and theirduration.Chance (1920) wrote .. . it is not known for how

long in a given case the influence of the alcohol canlast, nor has it been settled by post-mortem studies,I believe, just how great is the destruction of theganglia." So far as we can discover, these words arestill essentially true. Similarly, after the parallelprocedure of sensory rhizotomy, only a few patho-logical studies have been made of the sensory rootand central trigeminal pathway. Spiller and Frazier(1901) and others have made such studies in variousexperimental animals; but so far as we know, theonly three human cases that have been reported arethose of Sj6qvist (1938).

Cavina (1932) made an ingenious investigation ofthe site ofaction ofalcohol. On numerous occasionshe gave a trigeminal injection and immediately pro-ceeded to a posterior rhizotomy, having left theneedle in place; on opening the cave of Meckel henoted the position of the needle point. He was ableto infer that the alcohol had the greatest effect wheninjected among the fibres of the sensory root justanterior to the porus trigemini.We have discovered only two workers who

describe the histology of the trigeminal ganglion andthe adjacent trigeminal fibres after injection. Bothfound changes which might or might not have beendue to the alcohol, but neither attempted to deter-mine the site of the injection from the histologicalpicture. Guarch (1924) described two cases of ticdouloureux treated by alcoholic injection; theganglia were extirpated after an unspecified intervaland he stated that they were unaltered by the

* From tne Neurological Unit, Medical Research Council, NationalHospital, Queen Square, London.

injection, though he enumerated various chronicdegenerative changes. He noted chiefly that theganglion cells nearly all had dark staining nuclei,shrunken bodies and vacuolated cytoplasm; thecapsular cells were said to be phagocytic. In additionthe nerve fibres were grossly demyelinated. It isdifficult to see how he was so sure that the alcoholhad played no part in producing this degeneration.Heilmann (1929) also described a case in which atrigeminal injection had been made; the ganglioncells showed chronic degeneration and there wasdemyelination of both the central and peripheralportions of the nerve fibres. Neither of these paperscontains details ofthe degree of sensory disturbanceproduced by the injections. Without this informa-tion one cannot be sure that the alcohol came intocontact with any part of the trigeminal pathway.Proof that alcohol can destroy ganglion cells and

their fibres was provided by Chiasserini (1915).Using dogs, he injected alcohol under direct visioninto the spinal posterior root ganglia and examinedthem histologically after various intervals; alldegrees of cell and fibre destruction were found.As to the duration of the effects of alcoholic

injection on the human trigeminal nerve, the evidencefor this has hitherto been clinical (Harris, 1920;Carmichael and Woollard, 1933).The case described here offers some histological

evidence of the effects of alcohol on the humantrigeminal pathway and their probable duration.

NomenclatureThe term " Gasserian ganglion " has been used in

two different senses: (1) in the older sense, based ongross appearances, the ganglion is the swelling of thetrigeminal nerve and includes all the nervous struc-tures within the cave of Meckel; (2) in the newerhistological sense it is only the band of nerve cellsintervening between the three main divisions of thenerve and its sensory root. There is thus one struc-ture which belongs in the older nomenclature to theganglion and in the newer to the sensory root; thisis the aggregation of nerve-bundles between the

36

TRANS-SYNAPTIC DEGENERATION OF NEURONES

cellular band and the porus trigemini; it has longbeen called the plexus triangularis.We propose to follow Ferner (1949). He uses the

word ganglion in its newer and more restrictedsense, replaces the term plexus triangularis by theterm pars triangularis (of the sensory rqot), andcalls the remainder of the sensory root, from porustrigemini to pons, its pars compacta.

Case ReportIn June, 1945, a man, then aged 73 years, began to

suffer from increasingly severe pain which was alwaysconfined to the right cheek and forehead, occurred inparoxysms lasting only a few seconds, and was oftenprecipitated by eating, talking, washing the face orbodily exertion. In every way the pain was typical oftic douloureux. This continued until July, 1947, whenbecause of it he was admitted to hospital. On July 25,1947, a full neurological examination revealed nosignificant abnormalities. In particular the motor,reflex, and sensory functions of both trigeminal nerveswere normal. Routinp radiographs of the skull weretaken, also on July 25, and reported on by Dr. JamesBull as negative. The same day a right trigeminalinjection was given. When the needle point seemed tobe suitably placed, 0-1 ml. of procaine was injected;within one minute there was anesthesia to firm pin-pricks throughout the right trigeminal area. Next 0-5ml. of 90% alcohol was injected in divided doses of0-05 ml. The needle was left in place for half an hour,after which, there being no diminution of the anes-thesia, it was withdrawn. Because of certain abnorm-alities found during physical examination, radiographyof the chest had been performed. Dr. Bull reportedthat the film suggested the presence of large secondarycarcinomatous deposits in both lungs.The patient was re-examined at the second, third,

fourth, sixth, and eighth weeks after injection. Bylate October, 1947 he had begun to be cachectic, and onNovember 3 he was re-admitted to hospital. During thenext 12 days he was examined several more times, thelast occasion being on November 13. At every examina-tion there was anaesthesia to heavy pin-pricks throughoutthe right trigeminal area. The right side of the foreheadand the right cheek were tested for deep pressure sense;this was quite absent. On November 15 the patientdied.

Necropsy.-Twelve hours after death formol salinewas injected into the cisterna magna and the headrotated so that the fluid should run over the base andanterior aspects of the brain. The necropsy, unavoid-ably delayed until 60 hours after death, was performedby Dr. J. N. Cumings. The base of the brain and theemerging nerves were found to be well fixed. Thebrain and the contents of both caves of Meckel wereremoved. To the naked eye there was no abnormality ofthe brain, ganglia, or nerves and no secondary depositswere found, though carcinomatous masses were presentin the apices of both lungs. The other organs werehealthy.

Preparation of Sections.-Fixation was continued in20% formol saline for 14 days. The ganglia andattached nerve fibres were embedded by the celloidinmethod. The mid-brain, pons and medulla were slicedtransversely, alternate slices being prepared by ordinarycelloidin embedding and for staining by the Kulchitskymodification of the Weigert-Pal technique. Sections ofthe ganglia and fibres were cut serially at the angle whichgave the greatest longitudinal and transverse diameters.The stains used on the ganglia and fibres were Ehrlich'shematoxylin and eosin; Anderson's iron hmmatoxylinand van Gieson's stain; Mallory's phosphotungsticacid hematoxylin; thionin; Gros' silver stain; Loyez'myelin stain; and Gordon and Sweet's reticulin stain.Most of the sections were stained by Anderson's ironhiamatoxylin and van Gieson's method or by Gros'silver method, the other stains being used on representa-tive sections at different levels. As a control, six normalganglia and sensory roots were obtained at necropsies;five were embedded in paraffin and one in celloidin.These control specimens were sectioned as similarly aspossible to the injected one. The paraffin preparationswere cut serially, and, at intervals of 50, sections werestained by Ehrlich's haematoxylin and eosin, by Ander-son's iron hoematoxylin and van Gieson's stain, and bythionin. In the celloidin preparations every tenthsection was stained by Gros' silver method, and a fewsections by the himatoxylin and van Gieson's method.

Sections from the slices ofmid-brain pons and medulla,were stained by Kulchitsky's modification. of theWeigert-Pal method, Ehrlich's hiematoxylin and eosin,thionin, Loyez' myelin stain and Gros' silver stain.

Examination of SectionsFor descriptive purposes it is convenient to divide

the trigeminal pathway into a peripheral and acentral part. In all the figures (a) denotes normalcontrols, and (b) sections from the injected side inthis case.

Peripheral Trigeminal Pathway.-The sensory rootin the porus trigemini, a't its point of transition frompars triangularis to pars compacta, shows verydefinite changes. The usual formation of approxi-mately parallel bundles of nerve fibres is completelydisorganized. There are fewer bundles than usual;many are broken up into segments and are entangledin fibro-vascular tissue which is more abundant thanin the normal (Figs. 1 and 2). The very few remain-ing normal bundles lie near the supero-medialborder of the sensory root. Except in these fewbundles most of the nerve fibres show gross degener-ation, with almost complete loss of myelin; theSchwann cells are increased in number and there isa considerable leucocytic reaction (Fig. 3). The onlyaxis cylinders persisting are fine ones, which areswollen, and the argentophil material is mainlyin the form of granular detritus. The pars triangu-laris is very different from the normal. A few -finenerve fibre bundles run from the ganglion cells, but

37

JOHN PENMAN AND MARION C. SMITH

-'

n.

g.c. Tjaind. ri.

p-,/.,,.r1

FIG. la.-Section of normal trigeminalganglion and sensory root.

FiG. lb.-Section of injected trigeminalganglion and sensory root. (x 2.Anderson's iron htmatoxylin andvan Gieson's stain.)

o.n: ophthalmic division of fifth nerve;max.n: maxillary division; mand.n:mandibular division; p.t: pars tri-angularis of sensory root; por.t:porus trigemini; g.c: ganglion cells.

2a

FiG. 2a.-Area of porus trigemini innormal sensory root outlined inFIG. la showing bundles of nervefibres.

FIG. 2b.-Similar area in injected root,outlined in FIG. lb, showing dis-organized nerve fibres. (x 40. Ander-son's iron hlmatoxylin and vanGieson's stain.)

$

PLATE I

IV

38

v:1.

TRANS-SYNAPTIC DEGENERATION OF NEURONES

3b,-* F

S

.4 ff=s

- r e ' .

i96, Ja4,~'*p't4rtw;t~-1ta

s

t. 4.bw>*. ^

*-~~, J'w 4 * _S *0

~seb t ^a. * A *f.e 9 # -_ _ _ q. S L~~~~~~~~ sUt t>d* >

3a4s*|^* a s s;+ is so= Z "* t >o-.'*~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~I

¶

q

2%,

..

9- ..

it a.S

It. *6 a *

F1-1. 3a. Higher pover view of niorli-i1i nerve fibre 40r,tY* a

buLndle show-ni in F1i(. 2a. * A , < * t

Fic s. 3h anid 3(c.-Similar view-s ti-fom1 Fic;. 2b to show * £ , -Schwann cell itncirease. and ICLICOCNtic reaction. pP -1.(x; 320. Ati e so 's i oi liiiltoxdnanidvanG4ieson's-SL;lil1.> w : ; - ~~~~ w S

_ _ ~ ~~~ ~~~~~~-'_

FIo. 4.-Nerve fibre bundle near the ganglion cell zoneof the injected ganglion, showing large argentophilswellings. (x 200. Gros' silver stain.)

FIG. 5.-Similar area to FIG. 4.silver stain.)

PLATE II

39

1.~,ft

-~~--.-----.-.~~--. ---. - I,-----

;~~~~~ C=Atg $a s>~~ ~ * - bA;3Xitw

I * ISs ~'; ;i '* * s

4~~~~~~~~~~~~~~~4

t*t wsv; ~s;fNAp *v s_ *.ti, c |w

: j,t i~ * ^, W 94, *

v*

f*'iv.s; ~~~~~~*A,#i : .0.wi t b s .41F~~!

-> h 4 !F''*U1| \9 ¢ P ss

FIG. 6a.-Ganglion cells of left, normal trigeminal FIG. 6b.-Ganglionganglion. ganglion showinJ

( x 250. AndersGieson's stain.)

eT' W.cells of right, injected trigeminal

ig marked cellular proliferation.on's iron hLmatoxylin and van

Fio 7a- ion ells of left, normal ganglion Icondition of peric-61lular network and axis cy'Some thickenings of the processes are present

It

1,~~~~0

to show Fo. 7b.-Ganglion cells of right, injected ganglion

linders. showing many irregular large swellings on the axis

cylinders. (x 200. Gros' silver stain.)

82- - s1

Fios. 8 and 9.-Higher magnifivatiQn Qf pericellular and axis cylinder swellings as shown in FIG. 7b. (x 600.Gros' silver stain.)

PLATE III

9

f. ri.V111, "l

r,w.

~~~~~~~<.~~ ~ ~ ~~~~~vOI

4~~~~~~~~~t

A,I

fi11 1I a(i '"

FIG.I1 la.-Axis cylinders in intra-pontine par-fifth nerve.

Fia. 12a.-Main sensory nucleus of normal

rice I~~~loIt Oh. IrifLLtLd dI i flaII \cIV ist Tie eWIIr:JlMt( P0orts show 11W pa'oi du '10 ross`I`

b.p. b. o rind mio ion. 60I \Lm /' ,I lti sI-IaitIlIfibrves of, hrachluriri ponriti 1filt nli\ c.

$C~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~l

'00~ ~~~'

-,~~~~~~~~~~~A

.4h, Ma.~~~~~~~~~~~~~a

fifth nervel FIG. 1b.Mtain sensory nucleu of injected fifth nerve,

showing loss of nerve cells. There is marked microglialincrease. (x 80. Thionin stain.)

PLATE IV

FIG. 13.-Transverse section of medulla showing loss of axis cylinders inspinal tract of fifth nerve on right, injected side, compared with leftnormal side. ( x 7. Gros' silver stain.)

:i,11 - -''' wj _'*

FIG. 14a. Enlarged view of left, normal tract shown in FIG. 14b.-Similar view of right, injected tract, showingFIG. 13. loss of descending axis cylinders of spinal tract.

There is also a loss of secondary ascending axiscylinders in the hilum of the nucleus. (x 25.Gros' silver stain.)

d.p.a: descending primary axons of fifth nerve; s.n: region of cells of spinal nucleus; a.s.a: ascendingsecondary axons of fifth nerve; n.a: few normal axons; s.c: dorsal spino-cerebellar axons.

PLATE V

I

I

4

FIG. 15.-Transverse section of medulla at region ofpyramidal decussation. As in FIG. 12, loss ofdescending and of ascending axons in right, spinaltract of fifth nerve, seen. (x7. Gros' silver stain.)

a T

FIG. 16a.-Enlarged view of left normal tract. FIG. 16b.Similar viewofright injected tract, showingabsence of descending and ascending axons,as in FIG. 14b. No intact descending axonspresent at this level. (x 25. Gros' silver stain.)

- ~ ~ ~ ~ 150.4W *a.0 _J"A '.A-"

~~~~~~~~ '4~~~~~~~~~~~~~~~~~~~4

- b t½..wj,.~~~~~~~~~~~~~~~~~~~~~~~~~~~.

45 , _K_,,_,~ , t* .f ee

-- -,: .A.. l. --,9-,,, _ ; i-

=iti~~~~ ~~~~8 #*:'^ Sv .;, |te > >F^

a~~~~~~~~~~~~~~fOj< t ~~ nr_ - - a4*r.>~gtj,Jt_

t/~;!u'"asSt>r-eAc-/v;>t tIlF>>W tf,? flJr .. j_;- 3 2b*< fj0_>a;;;jf t'#'w'*l' <li!1b-'-' R0

AO~ ~ ~ ~ ~ a

'Sr rjt@4. ri,

* u |srH -[*~~~~~~~ ; jPSWSa*ebt4s16i:z"''-f= ' ffi / ^1 Wp

N-C

FIG. 17a.-Secondary ascending trigeminal tract on FIG. 17b.-Similar area on injected side, showing lossnormal side, at mid-medullary level, of fibres. ( x150. Weigert-Pal stain.)

PLATE VI

0~~~~~~~.f a* - 4

I~~~~~~~~~'s @i '4 '4

k~~~~. 1. .* , t.

S~ ~ S . .

. * I .t

* - _.

S.

~~~ ~ ~ ~ ~ ~ ~ ~~.

,#4*I

.S V e

FIG18 -Sia nulu of noma ftnre

S

4 -te t.

t~9

,.; r*U4.. .J

4I

4. I '-14 *

A .-

Ai . . 'A,*4:

I

0

w- f6'

*4b $'

V *44 ,

*i'.,.*

FIG. 18b.-Spinal nucleus of injected fifth nerve showingchromatolytic neurones. (x 200. Thionin stain.)

r... i

'A

9

* _

0

FIG. 19a.-Medium-sized neuronespinal nucleus.

.. - 3 ..-

.wuedAt'

..

w

Fj1;.. . X_ ...... ,# *F*_ X ..

.X.

:e:A; _i'};_

F __ .__.,.._ _

:^: :._ '_

S

p

*

I

r_r...

4

v

* Cs-cll

from normal FIGS. 19b and c.-Medium-sized neuronesfromtspinal nucleus of injected side,showing swollen cell bodies, eccentric orabsent nuclei, and chromatolysis.

PLATE VII

'pa

0_ ..

L.;IF. 4 a

L'Amm'IL_k

V

;0

-N. M,.4 .

TRANS-SYNAPTIC DEGENERATION OF NEURONESthey amount to only a fraction of the usual number(Fig. 1). The myelin staining of these fibres is moredefinite near the ganglion cells and fades off nearthe porus trigemini, as already described. There is agreat increase in cellularity. A few normal axiscylinders are present, but most of them are swollenand many end centrally in bulbous enlargements(Figs. 4 and 5). Fine silver-staining debris in clumpsappears to represent completely degenerated fibrebundles. Between the bundles is a fine fibro-vascularnetwork, continuous with the fibro-vascular tissuein the region of the porus trigemini. In some regionsthere seem to be spaces containing no tissue at all.Although the plane of the section may contributeto this appearance it must be partly due to actualloss of tissue, for the walls enclosing the spaceshave a smooth fibrous lining; small clumps ofargentophil material are seen round the spaces.This apparent loss of tissue is most marked near theinfero-lateral border in the superficial sections, andthe supero-medial border in the deep sections. It ispossible that necrotic tissue from the spaces mayhave escaped during the fixing and embedding. Theband of ganglion cells is of varying and irregularwidth exceeding the limits of irregularity seen in thenormal (Ferner, 1940). There is evidence of actualcell loss in the part of the ganglion cell band next tothe pars triangularis; here there are residualnodules, some still containing fragments of thedead neurones. In the hiematoxylin and van Giesonand in the thionin preparations there is a generalizedincrease in non-neuronal cellularity in the ganglioncell band, particularly marked on the side next to thepars triangularis. Many of the ganglion cells arenormal in appearance, within the limits of senilechanges (Ramon y Cajal, 1928; Opalski, 1930; deCastro, 1933; Truex, 1940; Andrew, 1941). Inscattered areas there is marked cellular proliferationand slight loss ofneurones (Fig. 6). The preparationsstained by Gros' silver method show thickenedpericellular networks and many irregular processesarising from cells; these processes have swollenbulbous' ends (Fig. 7). Similar structures arescattered among the cells (Figs. 8 and -9). Thedivisions of the nerve normally consist of nervebundles embedded in fibrous tissue; no increase inthis tissue can be seen. Some minor changes arepresent in the fibres of the ophthalmic and maxillarydivisions; these consist of occasional disintegrationof the sheaths, and ofvaricosities in the axis cylinderswithout break in the continuity. In the mandibulardivision these changes are rather more marked;here there is a greater loss of myelin, and bulbousswellings are present on the axis cylinders. Thechanges are not suggestive ofgross damage, and maybe, at least in part, artefacts.

Central Trigeminal Pathway.-The fifth nerve rootas it enters and traverses the pons is grossly disin-tegrated. The myelin sheaths are almost completelydisrupted, and the myelin staining material ispresent in irregular clumps and as granular detritus(Fig. 10). There are very few axis cylinders ofnormalappearance; the rest are represented only byargentophil debris (Fig. 11). The main sensorynucleus, consisting normally of rather small nervecells, exhibits here a definite reduction of the numberof the neurones; amongst the remainder many aredark staining and shrunken. The degeneratingfibre tracts show a marked microglial reaction, andin the sensory nucleus this increased cellularitymasks to some extent the loss ofnerve cells (Fig. 12).The degeneration seen in the intra-pontine part ofthe fifth nerve is present throughout the descendingtract, and is clearly seen in Gros', Loyez, andWeigert-Pal preparations (Figs. 13, 14, 15, and 16).In the medulla, that is, nearer the termination of thetract, the degeneration is absolute. The only fibreswhich remain intact at this level are a few situatedmost anteriorly; these must be the same fibres asthose which were found intact near the supero-medial border of the sensory root, and it is notimprobable that they belong to the ophthalmicdivision. Very definite changes are present in thespinal nucleus. In this nucleus the neurones are oftwo types, medium- and small-celled. About two-thirds of the medium-sized neurones are swollen,the nucleus is absent or displaced to one side of thecell, and there is moderate chromatolysis (Figs. 18and 19>. There is evidence of severe damage in a fewof the smaller neurones, which are shrunken withdark staining nuclei. Clumps of microglial cells areseen, suggesting neuronophagia of completelydegenerated neurohes.

Authorities are agreed that the secondary spinalfibres run in the hilum ofthe nucleus. On the injectedside in this case there is definite pallor of the secon-dary tract in both myelin and axis cylinder prepara-tions. The fibres of this tract run in small bundles,which in this case have lost many of their fibres(Figs. 13-17). No pallor could be detected where thefibres cross in the secondary tract; this, however, isnot surprising as the decussation is so gradual thatthe absence of a few fibres would not be demon-strable by the negative stains used.The mesencephalic nucleus and the motor nucleus

show no alteration from the normal.

Summary of Histological FindingsIn the trigeminal nerve the site ofmaximum tissue

destruction is the pars triangularis of the sensoryroot, where there is a great loss of fibre bundles, anddegenerative changes in the remaining fibres. Fibro-

45

JOHN PENMAN AND MARION C. SMITH

vascular tissue is normally present between thebundles; here it is much increased in amount.There is no extensive loss ofneurones in the ganglion.Some neurones adjacent to the pars triangularisappear to be completely destroyed; a few degenera-ting cells with marked non-neuronal cellular reactionare scattered among healthy neurones. There isthickening of peri-cellular plexuses; irregularclumps of argentophil material are present betweenand around the nerve cells, and generalized cellproliferation is seen. Many axis cylinders arisingfrom the neurones end centrally in bulbous swellings.Within the central nervous system, the main sensoryand the descending trigeminal tracts are degenerated,with loss of myelin sheaths and axis cylinders; theonly fibres surviving in these tracts are a few pre-sumably from the ophthalmic area. The mesen-cephalic nucleus is normal. There is evidence ofdegeneration in many of the nerve cells of the mainsensory and spinal nuclei, and considerable loss ofsecondary fibres. Very minor degrees of variationfrom the normal are present in the peripheraldivisions of the nerve.

DiscussionThe injection of alcohol has resulted in very gross,

probably irreversible degeneration of the nervebundles in the sensory root. It appears also to havecaused complete destruction of some ganglion cellsadjacent to the pars triangularis; probably thesewere the only cells directly affected by the alcohol.On the whole, the changes in the neurones, theirproduction of thickened processes, and any minoralterations in the peripheral divisions of the axonsare probably of a secondary nature, resulting fromdamage to the central parts of the axis cylinders.The injection of alcohol was made into the parstriangularis rather nearer to the pars compacta thanto the ganglion.The destruction or damage of the fibre bundles in

the sensory root has resulted in almost total degener-ation of the nerve tract within the pons and through-out its course as the spinal tract. The degenerativechanges of the secondary neurones in the mainsensory nucleus and in the spinal nucleus, and of alarge proportion ofsome fibres from these cells mostprobably indicate a trans-synaptic degeneration.After section of the sensory root (Spiller andFrazier, 1901 ; Sjoqvist, 1938), or section of thedescending spinal tract (Erskine and Rowbotham,1949), degeneration of the descending fibres of theprimary neurones has been described; so far as weknow, this is the first time that the secondaryneurones of the trigeminal tract have been shown toundergo trans-synaptic degeneration, though itsoccurrence in certain other tracts has been proven.The mesencephalic root is unaltered as would be

expected, for the mesencephalic cells are the primaryneurones (Carmichael and Woollard, 1933); anydamage to their centrifugal fibres in the sensoryroot need therefore cause no permanent changes inthese cells. Apart from these fibres and the veryfew others escaping unaffected at the time of in-jection, there is an irreversible loss of the intra-cerebral sensory part of the fifth nerve. Even if newcentripetal fibres were to grow from the neurones ofthe ganglion, they would be unable to extend beyondthe surface of the pons (Schafer, 1913 ; Spielmeyer,1922; Ramon y Cajal, 1928).

SummaryThe histological changes after an alcoholic tri-

geminal injection are reported. They consisted ofdegeneration of nerve fibres in the sensory root andthroughout the tract of the fifth cranial nerve.

Evidence is presented for a trans-synaptic degener-ation in the main sensory and spinal nuclei.

We wish to thank Dr. J. G. Greenfield and Dr. E. A.Carmichael for advice and criticism of this paper;Mr. Wylie McKissock for clinical facilities kindly pro-vided at his neurological unit; Dr. J. N. Cumings forhis kindness in performing the necropsy for us; Dr.James Bull for radiologica,l reports; and Dr. DenisWilliams for referring the patient to one of us forinjection. We also gratefully acknowledge our debt toMiss Vera Burgess and Miss Kathleen Harrison for theirtechnical assistance. The credit for the microphotographsis due to Mr. J. A. Mills to whonm we offer our sincerestthanks.

REFERENCESAndrew, W. (1941). J. Anat., Lond., 75,406.Carmichael, E. A., and Woollard, H. H. (1933). Brain,

56, 109.Castro, F. de (1932). " Cytology and Cellular Pathologyof the Nervous System," edited by W. Penfield.Hoeber, New York. Vol. I, p. 93.

Cavina, C. (1932). Riv. ital. Stomatol. ,1, 3.Chance, B. (1920). Arch. Ophthal., N. Y., 49, 621.Chiasserini, A. (1915). Policlinico, sez. chir., 22, 48,

96,,138, 175.Erskine, C. A., and Rowbotham, G. F. (1949). Arch.

Neurol. Psychiat., Chicago, 62, 493.Ferner, H. (1940). Z. Anat. EntwGesch., 110, 391.Guarch, F. S. (1924). Virchows Arch., 249, 495.Harris, W. (1920). Clinical Section, Proc. R. Soc. Med.,

13, 62.Heilmann, P. (1929). Virchows Arch., 272, 753.Opalski, A. (1930). Z. ges. Neurol. Psychiat., 124, 383.Ram6n y Cajal, S. (1928). 'Degeneration and Re-

generation oi the Nervous System." Oxford Uni-versity Press, London. Vol. II, p. 397 and 509.

Schafer, E. A. (1912). " Quain's Elements ofAnatomy",Eleventh Edition. Longmans and Green, London.Vol. II, Part I.

Sjoqvist, 0. (1938). Acta Psychiat., Kbh., SupplementNo. 17.

Spielmeyer, W. (1922). " Histopathologie des Nerven-systems." Springer, Berlin. P. 455.

Spiller, W. G., and Frazier, C. H. (1901). Philad. med.J., 8, 1039.

Truex, R. C. (1940). Amer. J. Path., 16,255.

46